In the near future, all of the 70,000-100,000 protein coding sequences in the human genome will become available, providing an invaluable resource for the discovery of new pharmacological targets. However, the realization of that value will require new methods for high-throughput identification and functional evaluation of new targets mined from the human genome. An important step in the functional evaluation of potential pharmacological targets is to determine with what other proteins they interact. Thus, there is much current interest in high-throughput mapping of pair-wise interactions among the members of the human proteome. However, current methods for detecting protein-protein interactions are not efficient enough for high-throughput work. We have developed a proprietary high-throughput cellular selection system, based on interaction-dependent beta-lactamase (Id(beta)-lac), which can produce selectable phenotypes directly from interactions of heterologous domains in the bacterial periplasm or eukaryotic cytoplasm. The goal of the present proposal is to adapt this system for rapid identification of as many pair- wise interactions as possible among single domains of the human proteome. Such interactions will be expected to provide a rich source of potential new targets for pharmaceutical intervention in disease. PROPOSED COMMERCIAL APPLICATIONS: The primary goal of this work is to develop new technologies to accelerate the process of identifying and evaluating new pharmaceutical targets from genomics data.